How do the oiling passages compare between the Cavalier and Ford?

 

To the best I can measure (sticking drill bits in the holes) the chevy, ford, and comp cams pushrods all have an oil hole with diameter between .094" and .109".

 

The 6.250 comp cams pushrods preload the lifter by turning the rocker screw one flat. That comes to about 60 degrees of rotation on M10-1.5 threads.
60/360*1.5 = .25mm (at the screw)
.25mm*2.6/1.6 = .406mm (at the lifter)
.406/25.4 = .016" of lifter preload

 

That's not enough, it will make your cam profile lazy. I would shim the pedestals and use the longer push-rods.

 

lazy? how so?

 

Not enough pre-load on hydraulic lifters is like having too much clearance on solids. You won't be taking full advantage of the cam, your engine will be noisy, and you run the risk of not only the lifter dropping hard on the trailing edge of the cam lobe, but pre-mature wear on cam lobes/lifters.

If you want to do this on a budget, I suggest you get the next size up of push-rods and shim the pedestals.

 

pilot bearing

 

the newly installed engine fired up on the first shot today. Practice makes perfect I guess. The valves did, in fact, tap slightly when it was cold. I will be ordering longer push rods and making some shims. There also seems to be a bad idler or tensioner bearing making a whining sound. Overall, no surprises yet.

 

Cross fingers and wait!

 

The .25 BAR (~4psi) spring is a tease. From experience, about how much boost can a stock 3400 handle, assuming a conservative AFR and timing?

 

We saw a few boost spikes in the low 20's when a WG reference line was damaged, but ran consistent high 12's in the 1/4 at ~12psi - 3400 with 3500 heads (9.1:1scr).

 

Thanks. I'm considering pulling some 3500 heads from the junk yard. In order to use 3500 heads, would I physically NEED 3500 head gaskets? Also, would I NEED the 3500 LIM?

 

Yes, and yes.

 

Actually, you can use the 3400 head gasket, but the LIM must match the heads. The 3400 gasket is the same thickness as 3500, the bore diameter changed, though, from 3.75" to 3.8" which will affect CR slightly.

Your CR with 3500 heads will be in the neighborhood of 9.1:1 which is a great one for boost.

 

Ok, good to know since I already have some new 3400 gaskets and KB/silv-o-lite hypereutectic pistons waiting.

The gasket bore is a difference of about:
pi*((3.80[in]/2)^2-(3.75[in]/2)^2)*.060[in]*2.54[cm/in]^3 = 0.3cc

 

3 hours and $117 later... excuse the cellphone camera picture. The 2006 malibu looked like it was rolled over and totaled.

 

I made this adapter plate out of 3/8" thick steel I had laying around instead of drilling and tapping the heads into the water jackets like I did last time. I had a friend mill 3/8" off of the back of the original cast aluminum bracket.

 

We have 200 miles on the 3.4/3500 with the 4psi wastegate spring. My brother did a 2nd gear pull against my Cobalt (24psi) today and he was still at my back bumper when we let out. It made me nervous for if/when we put the 8psi spring in the wastegate. We're still working out the fine tuning, but it has been a very fun car to drive with the t5 transmission.

 

Some informational notes...
1. Using a memcal from a 90-92 3.1L V6 F-body (memcal mounted knock board) the knock sensor is wired directly to pin B11 of the 1227165 ECM when using unmodified $6E code. The knock sensor is very useful for spark advance tuning when you log Knock Retard on a histogram that is set up to look exactly like the spark advance table.
2. Using a 3.4L rwd block, crank, and DIS parts, the default advance is 10 degress BTDC with the EST bypass connector unplugged. Therefore, the initial advance in the calibration should be set to 50 degrees BTDC so that the numbers in the Spark Advance table match the actual timing advance. I checked this to be accurate with a timing light.
3. A crude NPN-transistor inverting-level-shifter circuit makes the factory tachometer work with the DIS module.
4. Remeber to select the code 41 fueling mode in the calibration to match the number of cylinders and memcal. This was causing the ecm to jump in and out of single fire mode and do other weird things when it was not set properly. Five things must match for everything to work as intended: the number of cylinders in the engine, the cylinder select byte, the memcal, the code 41 fueling mode, and the DIS components.
5. It seems you can cheat a little bit of spark tuning resolution by changing the load scaling factor of the coolant correction table to 4 (yields 32-256 lv8 range instead of 48-160). The main spark table ends at 208 lv8, which the motor can exceed for most of the rpm range at only 4psi boost. The other method to get better resolution is to use bl85c's MAF/MAP $6E code and 3-bar MAP sensor.
6. A very useful data logging parameter for fuel tuning is what I called "Total Fuel Trim" (BLM+INT-128)/128. If you log this in a histogram against MAF voltage, multiplying the MAF calibration tables by the histogram averages will get the BLM within +/-5% on the first or second attempt.

 

Now you need to crank the pressure! Beware of tuning against false knock. If it's too rich (as turbo motors often are) it can be picked up as knock. Oil temp & pressure info will help distingush the 2.

 

How do you use oil temperature and pressure to identify false knock? I guess oil temp will be up if its too rich? The car is currently tuned to 12.3(+/-.1):1 AFR with WOT timing 16-18 degrees.

The 5.83 psi (red) wastegate spring was showing peak boost at 4psi on the gauge. The 8.7psi (blue) spring is showing peak boost at 5psi on the gauge. I'm beginning to wonder about the gauge or boost signal tube. I did a 40mph-75mph pull against a TPI350 corvette and I was about 2 cars ahead at the end. This motor pulls pretty good from 3500rpm untill 6500rpm when it starts to fade. Fun car.

 

Excessive timing and knock put alot of heat into the oil. When there's actual knock you can see oil pressure spike a bit on a datalog. I haven't witnessed oil temp climb myself (don't have it in the datastream), but according to grumpy (RIP) it's noticeable on datalog. Not really looking for evidence of false knock per se, you're looking for absence of evidence. Look for evidence of real knock correlating with knock counts on the log.

You usually won't see false knock until about 11:1 or so. I ran my car at 11.5:1 and occasionally it would peek into the high 10.X:1 range and it would show up as knock. An egt in the exhaust stream can also distinguish real knock, the temp will drop when it happens.

 

I checked the boost gauge againts a "precision" gauge at work and it checked out within better than 1/2 of a psi. I found a kink in the signal line too. I was logging max airflow around 180g/s with the 5.8 spring, now I see 220g/s max with the 8.7 spring.

I'm having a difficult time getting the idle AFR down to 14.7ish. The ecm is learning up to19-20:1 AFR (BLM128 and INT128) at idle only. One time the motor stalled at idle. I'm simulating a narrowband sensor with a wideband. All the other BLM cells are 124 and the Wideband reads high 14s-low15s during normal driving. Obviously, it dips in the 12's during PE.

 

I always had issues with my lc-1 and simulating NB.

 

Is it in open loop or is it "locked" to 128? I've noticed a problem with 6E on occasion where the ecm stops updating in certain BLM cells.

 

The closed loop bit logs true. I do catch it falling out of closed loop when the O2 sensor becomes "not ready". When I richen up the MAF table it will compensate the BLM untill the AFR is 19-20:1.

 

Ah yes, that's another thing with the lc-1, it won't register if voltage drops or the car is re-started and you have the power going through a "power off during start". I deal with this when tuning cars and plugging my portable lc-1 into a cig lighter and re-starting the car. You have to wait for the wb to reset.

 

I wired in a 4 wire narrow-band oxygen sensor a month ago. It fixed the idle learning issue. The stock replacement clutch started slipping after 2000 miles, so we slapped in a chinese 6-puck setup. This build has 2800 miles and counting. It feels like a 13 sec car at only 5psi, though the final goal is 8 psi. There is strong acceleration from 3500-6500 rpm and mid-20's mpg highway.

The build has changed a bit since the original plan years ago, so here's a recap of the current setup:
-3.4 RWD block, crank, rods, flywheel, and 2.8 balancer.
-cross-drilled main journal oil holes and drilled main bearings
-FWD windage tray and baffle removed from RWD pan
-HV/HP oil pump
-Hypereutectic 3400 pistons (rod/piston assembly match-balanced within 0.7 grams)
-3500 heads with yellow Z06 springs
-correct length pushrods (~.045" lifter preload)
-216i/216e .470"i/.470"e cam (really nice sounding idle)
-3400 UIM
-stock throttle body and accessories (with adapter brackets and modifications)
-GM DIS
-44lb injectors
-1227165 ecm with $6E mask tuning
-Turbo: T04E 60-trim compressor wheel in a .42AR housing / T3 turbine wheel in a .63AR housing
-fabricated "shorty-header" style manifolds with 1.68" ID
-35mm external wastegate
-full 3" exhaust with high-flow catalytic converter and resonator tips (no muffler).
-1994 v6 t-5 transmission

I still have all the parts laying around from the 3.1/3400 setup, but the block had a spun main bearing and the cranks both had a spun rod bearing. The 3100 pistons have like 300 miles on them and I have about 10 good connecting rods. I have the 3400 heads (dissassembled) and LIM. I'm not in a rush to get rid of anything, but I don't want them to go to waste if someone was needing something. I also have 6 used 3.4 pistons and 5 used 3400 pistons. I'll probably scrap all of it for cash eventually.